Enh: Support for scipy.sparse.sparray, the sparse array form in scipy.sparse

[dschult]

Please describe the purpose of the new feature or describe the problem to solve.

sparse currently supports __eq__ and __ne__ when the object is an instance of scipy.sparse.spmatrix. But it doesn't support when the object is scipy.sparse.sparray. As scipy.sparse moves toward the sparray interface it will be convenient to be able to do this comparison

Suggest a solution if possible.

The current code e.g. line 431 in sparse._umath.py checks isinstance with scipy.sparse.spmatrix.
That could check both spmatrix and sparray. I'm not sure where else this would need to be changed.

If you have tried alternatives, please describe them below.

No response

Additional information that may help us understand your needs.

No response

[hameerabbasi]

This has already been fixed in main, I just verified:

>>> import sparse
>>> import numpy as np
>>> import scipy.sparse as sps
>>> rng = np.random.default_rng(42)
>>> a_sps = sparse.random((100, 100), random_state=rng, format="csr")
>>> a_sps = sps.csr_array((a_sps.data, a_sps.indices, a_sps.indptr), shape=(100, 100))
>>> a_sps
<Compressed Sparse Row sparse array of dtype 'float64'
        with 100 stored elements and shape (100, 100)>
>>> a_sparse = sparse.random((100, 100), random_state=rng, format="csr")
>>> a_sparse == a_sps
<GCXS: shape=(100, 100), dtype=bool, nnz=198, fill_value=True, compressed_axes=(0,)>
>>> (a_sparse == a_sps).all()
np.False_
>>> sparse.__version__
'0.16.0a10.dev79+gd770f66'

Would a release (or backport) be desired?

[dschult]

Excellent!
And no release just for this feature is needed IMO.
Thanks